RU2504086C1 - Sychronisation network configuration - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: configuration of a synchronisation network node involves determining information on synchronisation sources of a plurality of synchronisation trails for passing synchronisation information from a synchronisation source to a node to provide a synchronisation reference signal. After determining automatically synchronisation transmission characteristics of trails which use packet-based communication, the trails are compared automatically, using their source information and their synchronisation transmission characteristics, for selecting which of these trails to use for providing the synchronisation reference signal for the node (N).

EFFECT: compared to a selection made based on source alone, using the synchronisation transmission characteristics of packet-based parts can enable a better choice of trail, and can enable comparison with synchronous type trails, and so enable hybrid synchronisation networks to be configured and maintained.

16 cl, 14 dwg

Description

FIELD OF THE INVENTION

This invention relates to methods for configuring nodes of a synchronization network, to nodes for such networks, to administration systems for such networks, to synchronization networks, to corresponding computer programs, and to databases for use in configuring such networks.

State of the art

There are many types of communication networks for transferring information between remote points. For example, global networks owned by telecommunication service providers, Internet service providers, enterprise intranets, cable television broadcasting systems, and other data transmission networks may use optical networks in which digital information is transmitted in the form of optical signals over optical fibers. Digital information in communication networks can be attributed to asynchronous or synchronous types. Synchronous types, such as SDH (synchronous digital hierarchy), require the exact operation of the general reference timing. This means that clock generators on one network node must operate at the same speed as clock generators on other network nodes.

To provide a common reference timing, communication networks carrying digital information may include synchronization networks to ensure that common reference timing is used throughout the communication network. One such synchronization network is described in a document of the European Telecommunications Standards Institute (ETSI) entitled European Guide (EG) 201 793 v1.1.1 (2000-10), entitled "Transmission and Multiplexing (TM); Synchronization Network Engineering". This document describes the various elements that make up the synchronization network, and the operating principles according to which such a network distributes accurate timing information from so-called primary reference clocks (PRCs) to clocks located in other pieces of equipment within the network. PRCs are the highest quality clocks in the network and are usually based on an independent cesium beam oscillator, providing very high accuracy of the clock signal.

Poor network synchronization usually leads to large amounts of jitter and drift and, consequently, to transmission errors and underload / buffer overflows. Both of these failures will lead to maintenance problems causing high error rates and unavailability of service. In the best case, poor synchronization creates only some inconvenience to some other network layer; in the worst case, it can cause the entire telecommunications network to stop carrying traffic. Thus, a well-planned synchronization network with support is a prerequisite for preventing or reducing the risk of critical failures in traffic networks.

Synchronization network planning is usually done manually according to some rules established in the relevant ITU-T recommendations (e.g. ITU-T G.803) and other relevant standards (e.g. the aforementioned ETSI EG 201 793). Some computerized tools can help plan and maintain a synchronization network, for example, by supporting an autonomous design for distributing timing reference signals, as well as providing simulation scenarios for normal operation and failure of a synchronization network.

On the other hand, the operational administration of the synchronization network is usually distributed across several platforms, since various types of equipment can be included in the synchronization network. Therefore, separate network administration systems must coexist, each of which takes care of the type of network, for example, dedicated synchronization, switching, transmission, etc. These administration systems usually provide only a means to monitor the synchronization network and identify possible failures; in the latter case, individual nodes will be reorganized according to their synchronization setting, or will be transferred to the operator for the implementation of restoration procedures.

In the case of a synchronization network based on the physical layer, the synchronization network is usually unallocated, in the sense that it is superimposed on the communication network, and the transport layer of this network is the carrier of the timing reference signal. For this reason, although this is the most common way, some problems may arise: this type of network is generally difficult to plan, difficult to operate, may depend on other operators and the type of communication network it is superimposed on.

Accurate planning of an unallocated synchronization network is, in general, a difficult task. Even if it is very well implemented in the initial planning, it requires considerable effort in re-planning the synchronization network with each change of other types / levels of the network.

Additional problems may arise in the development of new and diverse technologies. For example, synchronization networks may be hybrid networks consisting of chains (routes), some parts of which use synchronous transmissions and other parts of which use packet-based transmissions. For example, nodes supporting either TDM compatibility (e.g. PDH, SDH) or synchronous Ethernet can be mixed with those that use packet-based technologies to carry synchronization information. Administration of the synchronization network, in this case, looks very difficult and can significantly increase the OPEX (operating costs) of the operator.

Precision Time Protocol (PTP) (IEEE 1588-2008 Standard for a Precision Clock Synchronization Protocol for Networked Measurement and Control Systems) provides for sending timing information over packet networks, so Ethernet carriers, for example, can transfer to network nodes clock signals of telecommunication quality. This standard defines a structured timestamp based on the output of the master clock during transmission of the timing packet and its position in the Ethernet packet or IP packet. PTP uses a two-way transfer method so that slaves can generate frequency, time, and phase alignment. It defines leading nodes, boundary nodes, and transparent nodes. Packets can go through any Ethernet nodes and are considered regular data packets.

Disclosure of invention

An object of the invention is to provide an improved device or method for configuring a synchronization network. According to a first aspect, the invention provides:

a method for configuring a synchronization network node, the method comprising the steps of: a) determining information on synchronization sources of a plurality of synchronization circuits for transmitting synchronization information from a synchronization source to a node to provide a synchronization reference signal, wherein at least one of the circuits uses packet-based transmission for transmission of synchronization information, b) automatically determine the synchronization transmission characteristics of the part that uses the transmission based on pa etov, and c) automatically comparing circuit leading to the node, using the specific information source, and using the determined transmission timing characteristics to select which of these circuits and respective synchronization sources used to provide a reference clock signal for a node (N).

Compared to the selection made only on the basis of the source, by determining and using the transmission characteristics of the synchronization of the packet-based parts, it is possible to improve the selection of the circuit, and it is easier to compare the circuit with the circuitry that does not use packet-based transmission. This is especially useful for networks such as hybrid synchronization networks having different types of circuits containing packet-based parts and synchronous parts, for example, based on a physical layer connection. This can help to automatically configure and maintain such hybrid networks in a more uniform or universal manner. This reduces the time and money involved in the administration and maintenance of such hybrid synchronization networks. This is especially useful when communication networks are more complex mixtures of traditional TDM and packet-based technologies and have more widely distributed synchronization sources, which complicates their synchronization networks.

Any additional features can be added to those discussed above, and some of them are described in more detail below.

Another aspect of the invention may include a node for the synchronization network, the node having: a) a circuit information administration unit configured to determine information about the sources of the multiple synchronization circuits leading to the node, for passing synchronization information from the synchronization sources to the node, to provide a reference signal synchronization, moreover, at least one of the circuits uses packet-based communication to pass synchronization information, and the circuit administration unit the same is configured to automatically determine the synchronization transmission characteristics of the part that uses packet-based transmission, and b) a comparison unit configured to compare the circuits leading to the node, using the synchronization transmission characteristics to select which circuit to use to provide the reference signal synchronization for the node. You can add any additional features. Another aspect provides an administration system for a synchronization network, the administration system being configured to implement the above-described configuration method. This centralized configuration avoids changes in propagation delays over a large network and makes it easier for the operator to monitor or intervene.

Another aspect provides a synchronization network having a number of nodes and having a number of synchronization circuits for transmitting synchronization information from a synchronization source to nodes of a hybrid synchronization network, wherein the hybrid synchronization network has an administration system configured to implement the above-described configuration method.

Another aspect provides a computer program stored on a computer-readable medium and containing instructions that, when executed by the processor, instruct the processor to implement the above-described configuration method.

Another aspect provides a database for configuring a node of a synchronization network, the database comprising recording synchronization sources of a plurality of synchronization chains for transmitting synchronization information from synchronization sources to a node to provide a synchronization reference signal, wherein a part of at least one of the chains uses transmission to packet-based, and the database also contains a record of the transmission characteristics of the synchronization part using packet-based transmission, and recording which of the chains selected wound to provide a reference synchronization signal.

Any of the additional features can be combined with each other and combined with any of the aspects. Specialists in the art may see other advantages, especially over a different prior art. Numerous changes and modifications are permissible without departing from the scope of the claims of the present invention. Therefore, it should be clearly understood that the form of the present invention is purely illustrative and is not intended to limit the scope of the present invention.

Brief Description of the Drawings

The implementation of the present invention is described below by way of example with reference to the accompanying drawings, in which:

FIG. 1 is a diagram of a synchronization network,

FIG. 2 is a diagram of a synchronization network node according to an embodiment,

FIG. 3 illustrates some steps in configuring a synchronization network according to an embodiment,

FIG. 4 is a table of a source database and transmission characteristics for circuits according to an embodiment,

FIG. 5 to 9 are steps for configuring a synchronization network according to embodiments,

FIG. 10 is a diagram of a synchronization network node and a centralized synchronization administration system according to an embodiment,

FIG. 11 is a diagram of a synchronization network having two illustrative circuits leading to node N,

FIG. 12 and 13 are examples of sub-TLV synchronization specifications, and

FIG. 14 is a diagram of a communication network and a synchronization network.

The implementation of the invention

The present invention is described below in relation to specific embodiments and with reference to certain drawings, but the invention is not limited thereto, but only by the claims. The described drawings are only schematic and do not impose any restrictions. In the drawings, the dimensions of some elements may be exaggerated and not shown to scale for illustrative purposes.

Definitions

When the term “comprising” is used in the present description and claims, it does not exclude the possibility of other elements or steps. The use of nouns in the singular includes the plural of these nouns, unless otherwise specified.

The term “comprising” used in the claims should not be interpreted in the sense of limitation by the means listed below; it does not exclude the possibility of other elements or steps.

Elements or parts of the described nodes or networks may contain logic encoded on the media for any kind of information processing. The logic may comprise software encoded on a disk or other computer-readable medium and / or instructions encoded in a custom integrated circuit (ASIC), a user programmable gate array (FPGA), or other processor or hardware.

References to switching nodes can cover a switching node of any kind, without being limited to the types described.

Software references can encompass any type of program in any language, executed directly or indirectly on processing equipment.

References to hardware, equipment, or processing circuits can cover any kind of logic or analog circuits with any degree of integration and without limitation, general-purpose processors, digital signal processors, ASIC, FPGA, discrete components or discrete logic, etc.

Links to chains can cover any indication or description of the path traversed by synchronization information, such as a list of nodes or a list of connections between nodes, or a list of locations or directions of passage, or an algorithm for computing such lists or directions, or any other similar indication or description.

Links to network nodes can cover any kind of identifiable network location that has any function, such as routing or switching, or multiplexing, or demultiplexing, or to support various transport technologies, for example, OTN, SDH, or some other processing of information passing over the network, without limitation by one or another level of integration, or by size, or band, or bit rate, etc.

References to a packet basis can encompass any kind of packet that has a header or frame of any kind that can be transmitted asynchronously or transmitted over a synchronous connection. In relation to synchronization, such packets may have a timestamp, for example, in the header or payload. Examples are NTP packets or PTP packets defined in RFC 1305, and many other types can be envisaged.

References to hybrid synchronization networks may include networks in which timing information (e.g., frequency or phase) is transferred in some parts at the physical level and in other parts is transferred through packet-based transmission, unlike networks that do not use packets or networks in which all information is transported by packets, even if packets are transported at the synchronous physical layer.

References to timing sources can cover any timing source in the timing chain and can cover PRC for frequency, and primary timing reference signals for PRTC, for example, for phase information. This can be realized, for example, in GPS (global positioning system) or in atomic clocks.

These definitions can be expanded when appropriate to cover definitions of similar parts in standards such as the well-known G.810.

References to synchronization transmission characteristics can cover any kind of characteristics that affect the transmission of synchronization information, for example, the number of nodes passed, the speed of connections used, the status of parts or connections used, quality characteristics, such as packet delay changes, nodes that regenerate synchronization information, nodes, which expand synchronization information, for example, simultaneously taking measurements.

Abbreviations

ASON - automatically switched optical network;

BMC is the best leading clock;

ESMC - Ethernet sync messaging channel;

LC is a component of the compound;

LMP - connection administration protocol;

LSA - connection status notification;

MPLS - multi-protocol label switching;

OPS - route support;

OSPF - the first discovery of the shortest routes;

PDV - packet delay change;

PTP - exact time protocol;

QoS - quality of service;

SSM - synchronization status message;

SyncE - synchronous Ethernet;

TE - traffic engineering;

TLV - type-length-value;

WSON is an optical network with wavelength switching.

Anticipating the description of the embodiments, we will consider some of the problems associated with the traditional features of synchronization networks in order to facilitate understanding of the features of the embodiments and how they combine or contrast with the features of existing networks.

FIG. 14. Communication network and synchronization network

In FIG. 14 is a block diagram of a digital communications network 100 that includes a synchronization network. For purposes of illustration, network 100 is a telecommunication network and therefore includes, at network nodes, equipment well known in the art. In the figure, transport connections are indicated by solid lines, and synchronization circuits for transferring synchronization information (circuits formed by a sequence of reference connections) are indicated by dashed lines that include an arrow at one end to indicate the source and recipient of synchronization information, for example, a reference clock signal. When a node has the ability to receive a reference clock signal from more than one source, primary reference connections (ie, synchronization connections that are preferably used to supply a reference clock signal from one node to another) are indicated by the number "1" after the dashed line indicating the connection . Secondary reference connections (ie, synchronization connections that are used when the primary synchronization connection is unavailable) are indicated by the number "2" after the dashed line indicating the connection.

When some of the backbones that make up the synchronization chains are designed to use packet-based protocols, the synchronization network is called a hybrid synchronization network. Such packet-based protocols can be carried over synchronous physical layer connections of the underlying transport network, as shown in FIG. 14. An example in more detail which connections use packet-based protocols is shown in FIG. one.

Network 100 uses a synchronous digital hierarchy (SDH), which is the standard technology for synchronous data transmission over optical communication media. It is the international equivalent of Synchronous Optical Network (SONET). To facilitate the following consideration, the various nodes of the network are equipped with the symbols A, B, C, D, E, F, G, H, I, L, M and N.

In a fully synchronized network, all sources must be finally traced to the PRC. In an illustrative network, this is PRC A. PRC A feeds its high quality clock (“clock”) to node B, which may be, for example, stand-alone synchronization equipment (SASE) B. SASE is an instance of synchronization equipment that contains a block Synchronization Support (SSU), which is a high-quality subordinate clock. SASE B distributes its clock signal to node C, which may be, for example, a digital switch (or, in alternative embodiments, may be a telephone switch). Node B can also provide a signal to node D, which can be, for example, an SDH (MUX) D multiplexer.

The SDH D multiplexer distributes its clock signal to the SDH digital cross-connect unit (DXC SDH) E, which, in turn, distributes its clock signal to the (channel selective multiplexer) Add Drop Multiplexer (ADM) SDH F. The clock signal output ADM SDH F then goes to each of two or more ADM SDH Gs and I. The support connection between the ADM SDH F and ADM SDH G is the primary connection.

Instead of using the supplied clock signal itself, the ADM SDH I operates in “bypass” mode (usually referred to as “NON-SETS locked”, where “SETS” is the abbreviation for “synchronous timing equipment source”), so that the synchronization clock signal is directly redirected only at SASE L. ADM and SASE, for example, are usually implemented in the same building. In essence, SASE L is the actual recipient of the synchronization clock provided by the ADM SDH F, and thus, the clock is regarded as a secondary connection. The primary SASE L connection is provided (via ADM SDH I operating in "bypass" mode) via ADM SDH H.

Despite its bypass function, the ADM SDH I requires a clock clock, and this is achieved through SASE L.

The ADM SDH I feeds its clock clock to the ADM SDH H, and this is considered a secondary connection. The primary ADM SDH H connection is provided through the ADM SDH G. To ensure reconfigurability, the ADM SDH H is also connected to provide a clock clock to the ADM SDH G, and this is considered as a secondary connection via the ADM SDH G.

The ADM SDH H also supplies a clock signal to digital switch M, which also receives a clock signal from digital switch N. The rest of the network is not shown for brevity.

FIG. 14. Timing Cycles

It is very important to plan the synchronization network in such a way as to avoid the occurrence of timing cycles, both during normal operation and when a failure prevents one or more nodes from supplying their reference clock signals to the corresponding scheduled receiving nodes. The timing cycle occurs when the clock signal is directly or indirectly synchronized with itself. In a timing cycle situation, all clock signals belonging to the cycle may exhibit a large frequency shift relative to the rated frequency and are very likely to be isolated from the rest of the synchronization network. To avoid timing cycles, the elements in the ring should be equipped with a means to detect the possible generation of timing cycles. Such elements are usually connected so that each of them has at least two sources of synchronization, due to which, in the case of identifying one source that generates a timing cycle, at least it is possible to avoid it by choosing one of the alternative sources. For example, suppose that the support connection between nodes F and G is broken. In this situation, the ADM SDH G will in response orient itself to the H node to provide the necessary reference clock signal. However, under normal circumstances, node H expects to receive its reference clock from node G. Obviously, a timing cycle will occur here if node H also does not respond to a disconnection between nodes F and G, focusing on a different source with respect to its reference clock . It is important that the clock provided by this alternative source is also not permanently outputted from the clock at node G or from the clock at node H in order to avoid a timing cycle.

On SDH networks, the use of synchronous status messages (SSMs) helps to some extent to avoid timing cycles. SSM is a signal that is transmitted through the synchronization interface to indicate the quality level of the clock generator, the interface to which is finally traced; that is, the quality of a clock generator with which it is directly or indirectly synchronized through a chain of clock generators of network elements (a "synchronization chain"), no matter how long this chain of clock generators is. In a fully synchronized network, all sources must be finally traced on the PRC, and a predefined code exists to indicate this. Another code, “Do not use for synchronization,” is used to prevent timing cycles and is transmitted in the opposite direction through the interfaces used to synchronize the equipment clock.

Although the SSM algorithm is a convenient approach in a number of applications like SDH or SONET rings, it is not able to guarantee the prevention of all timing cycles, since it provides only information about the quality of the trace source of the synchronization reference signal, but not information about the actual physical source. See, for example, chapter 4.13 of ETS 300 417-6-1, "Generic requirements of transport functionality of equipment: Synchronization layer function". Another drawback of the SSM algorithm is that it is often not supported by SASE or network elements other than SDH / SONET network elements (i.e. it can only be used between SDH / SONET network elements).

It should be noted that timing cycles can cause serious violations in the traffic network, however, the result of these violations very rarely gives an easily distinguishable indication of where the failure in the synchronization network occurred. Therefore, it is important to provide effective ways to administer synchronization networks so that when network failures occur, you can determine how to reorganize the network to maintain an acceptable quality of synchronization without creating timing cycles.

FIG. 14. Configuration Administration Issues

Current practice involves the distribution of the administration of synchronization networks between several platforms. The reason for this is that the synchronization network very often consists of various types of equipment that can be designed either for synchronization (for example, SASE), or for synchronization and traffic (for example, an SDH multiplexer or digital switch). As a result, it is necessary to simultaneously support several administration systems (for example, one for the SASE network and one for the SDH equipment, one for the switched network, etc.). This situation is shown in FIG. 14, where the first dial-up network administration network 101 controls a digital switch at node C; the SASE administration network 103 controls the SASE at nodes B and L; the first SDH administration network 105 controls the SDH multiplexer on the Node D, the DXC SDH on the Node E and the ADM SDH on the Nodes F and I; a second dial-up network administration network 107 controls the digital switches at nodes M and N; and the second SDH administration network 109 controls the ADM SDH on nodes G and H. This situation is not unrealistic because the "same" type of equipment (eg, ADM SDH) can be manufactured by different manufacturers who design their equipment using incompatible strategies equipment administration. Some or all administration systems can be configured to support a local map of neighboring nodes and various circuits providing synchronization information from various sources to each of its nodes. These administration systems must be individually configured with details of any changes outside their management area.

When replacing the method of manipulating synchronization information on some nodes to use packet-based transmission, the synchronization network turns into a hybrid network. The network, where the timing information is transferred in the form of packets, allows the distribution of phase and time information, as well as frequency information. When some parts of the circuits use packet-based transmission, this can further complicate the administration of the synchronization network configuration, both during initial planning or at the commissioning stage, and during further configuration during the life of the networks. This may include continuous service from the staff, which may require its direct presence at numerous locations of various nodes. In most cases, this direct presence at each location is difficult and involves financial costs due to the geographical distribution of the synchronization network.

Features of Embodiments

To solve the above problems and to reduce the associated operational costs, one approach involves trying to add a certain level of automation to the configuration of the synchronization network, from initial planning to reconfiguration during operation of the synchronization network. A more decisive approach involves the introduction of a control plane of the communication network to control the configuration and operation of the synchronization network. This is one of the features of some embodiments of the present invention. Other features for expanding traditional configurations include a fully automated synchronization network implemented by the control plane, where, in addition to a powerful synchronization operation (for example, failure recovery), the synchronization network can be independently configured, starting with the topology and other relevant information.

In addition, the concepts can be applied to any type of synchronization methods, not only to SDH / PDH, but also to new generation technologies, such as IEEE 1588 based synchronization networks. In addition, some embodiments may include time synchronization in addition to frequency synchronization, and some embodiments may use OSPF extensions to determine the LSA of a TE connection, which can also carry synchronization information.

Some embodiments may have distributed administration of the synchronization network configuration, in which all nodes participating in the synchronization network exchange information to build a picture of the synchronization network itself and its synchronization chains. This can be used to maintain it (for example, to select a synchronization reference signal according to predefined rules) and to enable the corresponding response by reconfiguration in the event of any change in topology or network failure.

This can be done in a centralized or distributed mode: in the first case, the central network administration node assumes the function of distributing all the necessary information and configuration commands. In the latter case, each node maintains a database containing all information related to the entire network.

As described in more detail below, the configuration of a hybrid synchronization network node may include determining information about the synchronization sources of the multiple synchronization circuits to pass synchronization information from the synchronization source to the node to provide a synchronization reference signal. After automatically determining the synchronization transmission characteristics of the chains that use packet-based transmission, the chains are automatically compared, using their source information and their synchronization transmission characteristics, to select which of these chains to use to provide the synchronization reference signal for the node. Compared to source-only selections, the use of packet-based parts synchronization transmission characteristics can provide better circuit selection and can provide comparisons with synchronous-type circuits, which can help to configure and maintain such hybrid synchronization networks.

Additional features of some embodiments may include the following: the step of determining transmission characteristics of the synchronization may comprise determining information about nodes used to transmit packets used for transmission based on the packets.

This characteristic probably has the greatest impact on the quality of synchronization transmission through the packet.

The step of determining node information may include determining information related to the expansion of synchronization information carried by packets by any node. Such an extension can improve the quality of the chain and, thus, can affect the comparison of the chains and improve it. The step of determining information related to the extension may include determining information related to the regeneration of synchronization information on another node (F) along the chain. Such regeneration can improve the quality of the chain and, thus, can affect the comparison of chains and improve it.

The extension may contain delay information for packets carrying synchronization information, the delays being determined by another node (H) along the chain. The node can use such additional information to improve the quality of the reference signal, and this can affect the quality of the circuit and, thus, can affect the comparison of circuits and improve it.

The delay information may contain any one or more of: delays internal to the node, delays between nodes and delay variation values. The method may have an additional step of determining information about the synchronous parts of any circuits leading to the node and using this information in the comparison.

The step of determining the characteristics of the synchronization transmission can be carried out by the node, and the node can have a record of the synchronization chains with itself and with other nodes. Such distributed configuration management can increase local or central equipment fault tolerance and facilitate the addition of new nodes, although centralized solutions may be better suited for larger networks to prevent slow propagation of changes.

The method may have the steps of receiving on the update node the synchronization transmission characteristics from the neighboring node, reviewing the selection based on the updated synchronization transmission characteristics, selecting a new one of the synchronization circuits to pass synchronization information to use as the synchronization reference signal, and sending updates to neighboring nodes to indicate Updated synchronization transmission features and a new circuit selection. By including the current reconfiguration, as well as the initial setup, you can achieve the maximum reduction in operating costs.

The synchronization transmission characteristics determined by the circuit administration unit may contain information related to the expansion of synchronization information carried by packets by any node.

A node may have a part of a control plane for controlling a communication network synchronized by a synchronization network, wherein the network administration unit and the comparison unit are implemented as part of the control plane. This can be convenient, since such a control plane infrastructure already exists and has the ability to observe the topology and changes that can affect the synchronization network configuration.

Some of the described embodiments can significantly reduce operating costs when deploying next-generation hybrid synchronization networks. This can be achieved by setting a new powerful mechanism integrated into the control plane and able to work with various types of synchronization networks (both during configuration and during network operation).

They can be based on mixed methods based on packets and on the basis of the physical layer and can in some cases provide frequency synchronization or time synchronization (or both). We list some of the characteristics described in more detail in the following section:

distributed synchronization network administration;

extension of automatic manipulation of the synchronization network to methods on a mixed basis of the physical layer / packets;

extending automatic manipulation of the synchronization network to methods on a mixed basis of the physical layer / packets to support time synchronization.

Although the distributed approach will be described first, the initial network setup can be determined by the operator. Any reconfiguration during operation may be carried out autonomously or, alternatively, automation may also extend to part or all of the initial setup of the synchronization network.

This is similar to that defined in the aforementioned IEEE 1588 best lead clock algorithm in the field of packet-based synchronization: a signaling channel (alert message) is used to communicate between all IEEE 1588 nodes and, for each node, to specify its role in the network and, for slave clocks, to specify their master clock.

FIG. 1. An example of an automated synchronization network

According to FIG. 1, the synchronization network 200 contains nodes designated A, B, C, D, E, F, G, H, I, L, M, and N, which are nominally connected to each other, forming a configuration similar to the synchronization network 100 shown in FIG. . 14. The synchronization network 200 is a hybrid network since some parts of the circuits use packet-based transmission. This is an example of a network that has distributed configuration administration, since it has the ability to route synchronization connection status information within the synchronization network 200 on each node, which allows each node in the synchronization network 200 to maintain complete dynamic connection status information of the synchronization network 200 at any point on the network. In this illustrative embodiment, the synchronization network 200 further includes an integrated synchronization network administration network 201, which is capable of administering all nodes in the synchronization network. The provision of a centralized network 201 for integrated administration of the synchronization network is not an essential feature of all embodiments, since, as explained above, this function can alternatively be distributed between different nodes, each of which has complete information about the circuits of the synchronization network 200 and, in particular, their dynamic connection status.

According to FIG. 1, each node in the synchronization administration network is assigned an address (for example, an IP address) that uniquely identifies this node. For convenience, the conventions A, B, C, D, E, F, G, H, I, L, M, N, and O used here to designate nodes will also be used here to represent the corresponding addresses of these nodes. Each node further has a storage device for storing a table that defines a relationship between each node and other physical connections in the synchronization network 200. In FIG. 1, this is schematically represented by several configuration information schemes 203. Some of the nodes are configured to use packet-based communications for parts of circuits. The connections shown in this example as using packet-based transmission are indicated by the following pairs of nodes EF, FG, GH, HM, MN, OF, FI, IH.

FIG. 2-3. Distributed circuit selection

According to embodiments involving configuration administration distributed between nodes, equipment at each node has an interface capable of supporting IP, and all this equipment is connected to an IP network. However, it should be emphasized that the use of IP standards is not an essential feature, and that other standard telecommunication protocols can be used instead, provided that these other protocols allow the transfer of information between nodes, described below.

In some embodiments, each node additionally has a storage device for storing a table that defines a relationship between each node and other physical connections in the synchronization network 100. In FIG. 1, this is schematically represented by several configuration information schemes 203.

In FIG. 2 shows a diagram of an illustrative assembly. She shows the node 400 of the communication network. It comprises a synchronization network node 410 and other features not used by the synchronization network, for example, a synchronous transmission processor 470. It serves to process synchronous traffic on transport connections of a communication network according to traditional practice. It is synchronized by a synchronization reference signal or a clock signal CK output by the synchronization network node 410.

This node of the synchronization network has a circuit selection unit 440 for selecting which of the various synchronization chains leading to this node is selected to provide synchronization information. This synchronization information is used by the synchronization reference signal generator 450 in the usual manner to generate a clock signal. The circuit selection unit is controlled by the output of the circuit comparison unit 430. This output signal effectively controls the configuration of this part of the synchronization network, controlling the choice of particular circuits.

The circuit comparison unit uses information about the source and the characteristics of the synchronization transmission stored in the database 435. The circuit information administration unit 420 fills and maintains this database by exchanging information about neighboring circuits with neighboring nodes, which is indicated in the figure as “circuit information message”. In practice, this circuit information about the sources and characteristics of the synchronization transmission circuit can be carried along the same physical routes that are used by the communication network, or used by the synchronization chains themselves, or using independent routes. Similarly, synchronization circuits can use the same physical routes that are used for communication network traffic, or have independent routes. The circuit information administration unit and the circuit comparison unit can be implemented in the form of software modules executed, for example, by a general-purpose processor, or, if desired, can be implemented in the form of software executed by separate equipment. The node may have a part of the control plane 425 according to traditional practice, distributed over all nodes of the communication network, for distributed traffic control in the communication network. The same control plane can be adapted and used to implement the functions of the circuit information administration unit and the circuit comparison unit.

FIG. 3. Stages of work when configuring synchronization circuits

First, consider the steps described below as a whole. The tables on each node are initialized to express the initial state of the synchronization network. This can be done according to any communication protocol between nodes, as discussed above. Each node (including the integrated synchronization network administration node, if one is provided for by the embodiment) now has a complete picture of the chains indicating how the reference clock signals propagate within the network. Then, the synchronization network is operated, using well-known techniques, to distribute the reference clock synchronization signals to each node along the chains. Given the absence of problems or other changes, this continues indefinitely.

However, whenever there is a change in status on a node, the table on the changed node is updated. Then, the communication protocol between the nodes is used to distribute this information to all other nodes in the synchronization network, including the integrated synchronization network administration node, if one is provided for by the embodiment. When disseminating updated information within the synchronization network, one or more nodes may respond with their own status / configuration changes, which requires additional table updates. Thus, by means of well-known iterative methods and a communication protocol between nodes, additionally updated information can be distributed to all other nodes in the synchronization network, including the integrated synchronization network administration node, if it is provided for by an embodiment.

In the end, no additional changes are made to any of the nodes, and each node, again, has a complete picture of how reference clock signals propagate within the network. Then, the synchronization network can be exploited again, using well-known techniques, for distributing the reference synchronization clock signals to each node. The above approaches make it possible to fully implement all the synchronization network administration functions in a uniform manner, nevertheless distributed between nodes, or under the control of one (centralized) node. In this regard, it is especially important that new types of equipment (in particular IP routers) also begin to need to manage the synchronization network administration network.

In FIG. Figure 3 shows some of the steps for configuring a node either through a centralized administration system or in distributed mode on each node. At step 10, a determination is made which chains lead to the node. At step 20, source information of each circuit is determined. At step 30, information about the transmission characteristics of the synchronization of parts based on the packets of each circuit is determined. The definition of the synchronization circuit information at each node may include the transfer between nodes of information regarding the synchronization chains and status, or may include the search for such information previously stored in the database. The different circuits are compared in step 40 using at least the appropriate synchronization transmission characteristics and source information and, optionally, other circuit selection information used to provide the synchronization reference signal for the node. These steps can, if desired, be implemented as hardware or software executed by a general purpose processor. Examples of the implementation of each of these stages of determination and comparison are described in more detail below.

Comparison and selection of circuits based on OSPF principles

Based on this information of the circuits, a node (any network element) can specify alternative sources of synchronization without the danger of creating timing cycles and in accordance with the network synchronization plan. Alternatively, as mentioned above, the process may use some more advanced mechanisms, where the network element selects the most convenient synchronization reference signal based on some suitable algorithm (for example, minimizing the chain length). An output signal indicating which circuit to select is sent to control the circuit selection unit on the assembly. In this context, although not shown, the selection unit can, of course, include an internal circuit from the internal clock signal generator, which, of course, should not pass through any other nodes, or through any connections between nodes, in in this case, the chain can be represented simply by indicating that the chain is internal. An illustrative process for making this choice of circuitry can use the principles applied in the OSPF (Priority Shortest Path Initialization) protocol, as shown, for example, in US 6711411 B1 “Management of Synchronization Network” - Stefano Ruffini, March 23, 2004. OSPF can provide fast response in in case of changes in network topology and flexibility in the reorganization of the network based on the appropriate optimization algorithm. Thanks to the extension of the OSPF protocol, all nodes in the synchronization network have information about the synchronization connection status of the entire synchronization network.

The basic principles apply to a network synchronization application. It is assumed that each node must be assigned a unique address.

The automatic determination of information about synchronization network circuits from other nodes can use principles known in connection with the routing protocols of the connection state, for example, the data routing protocol called “Open Shortest Routes Priority” (OSPF). These principles are preferably applied in the present invention with respect to the administration of the synchronization network, since they provide a quick response in case of changes in the network topology, low overhead traffic and flexibility in the reorganization of the network based on appropriate optimization algorithms.

This should not be understood in the sense that traditional connection state routing protocols are always useful for the administration of a synchronization network - this is not so if, for example, it does not provide data exchange related to the administration functions of a synchronization network (for example, data related to synchronization chains and status). However, the OSPF routing protocol allows all routers on an IP network to have a complete picture of the IP network so that each packet of data can be routed from its source host to its designated destination host, and now we describe how these concepts can be used to make it easier to use Network administration synchronization.

In OSPF, all routers are provided with a complete picture of the IP network due to the fact that each router sends updates of its connection status (i.e. information about changes in the status of the established connection) to other, neighboring routers in the IP network. Recursively, the same information is distributed within the network until each router has information and can not make a complete picture of the IP network, which can be used when routing IP packets.

Messages used for inter-nodal exchange of circuit information

In accordance with at least some embodiments of the invention, a similar approach can be used to provide each node in the synchronization network with complete information about the connection status of all other nodes in the synchronization network. This requires information from the active synchronization reference signal and its status (i.e., the identification of the traced synchronization source and its quality; and the identification of the pending (backup) synchronization reference signals and their status).

A number of messages are required to distribute information between nodes. Similarly, the following message types defined in OSPF are required:

(synchronization) greeting - for periodically informing connected nodes of the synchronization status of the sender;

(synchronization) database description - for informing about active and waiting synchronization reference signals and their quality status in the initialization phase;

(synchronization) connection status request - to update synchronization data, for example, after receiving a database description by a node and detecting its own database obsolescence;

(synchronization) update of the connection status - to inform other nodes about changes in the synchronization status;

(synchronization) acknowledgment of the connection status - to acknowledge receipt of an update of the connection status.

Note: all of these messages are sent to neighboring nodes. All nodes in the network, however, recursively receive information.

If messages come from nodes that are not adjacent (according to the topological information available to the receiving node), they are ignored. A list of neighboring nodes is generated via (synchronization) hello messages. When setting up the network, this can also be set by the network administration node to speed up the configuration process. Additional messages are required to prepare and configure the network, for example:

a command indicating which reference signal should be considered an assumed synchronization reference signal;

a command that prioritizes the various reference signals;

one or more commands that set thresholds for control parameters, if any. The sequence of steps in administering such a network can be briefly indicated as follows:

nodes are configured from a centralized system using some basic information (for example, priorities of the leading clock generators and, in general, information about the topology of the synchronization network specified by the operator);

the tables on each node are updated using message passing (synchronization) greetings and (synchronization) database descriptions). This corresponds to the above steps 20 and 30 in FIG. 3. Each node (or, in some cases, only a centralized node) has a complete picture of the synchronization network;

the network is operated accordingly;

when a change occurs on node X, the table of this node is updated (again, in accordance with steps 20 and 30), the chain selection for this node can be reviewed (in accordance with step 40 in Fig. 3), and the new information extends to all other nodes in the network (including a centralized node) by sending a message (synchronization) to update the connection status.

When disseminating information within the network, one or more nodes can respond with their own changes, thus requiring additional changes in the table of node X. The iterative process will eventually lead to a new status of the synchronization network.

The example shown in FIG. 5 having some similar steps. In the case of a fully automated network, the initial configuration may include only basic data (for example, the name and location of the lead (s) clock (s) generator (s)) and other restrictions set by the operator. The selection of the circuit (also called the choice of the reference synchronization signal) made by each node can be carried out according to some iterative process (similar to the case of recovery after network failures) and based on some suitable rules and algorithms.

FIG. 4. packet format and state table

In FIG. Figure 4 shows an example of database information that each node has at some point in time. This complete table can be sent in messages (synchronization) database descriptions to neighboring nodes. A (synchronization) update of the connection state usually includes only a portion of the table requested in the (synchronization) query of the connection state (for example, rows related to a particular node). Packets (synchronization) greetings may include information only about the status of the sender (for example, the status of established synchronization connections).

In this example, the table has several rows for each of the nodes shown in FIG. 1, with each row corresponding to a separate chain leading to each node, or an internal chain for this node. In the first column 301 of the table, each of the nodes in the synchronization network 200 is defined. For each of these nodes, the entry in the second column 303 of the table identifies the source of the node synchronization. For example, node A is a PRC (see FIGS. 1 and 14) and thus has no other sources. Node B has a connection for receiving synchronization from node A, and this is indicated in the second column 303. In addition, for node B it is the preferred source of synchronization, therefore, in the third column 305 of the table there is an indication that this source is assigned the highest priority (for example, priority "one").

Many nodes have more than one possible source of synchronization. For example, node B has a connection for receiving synchronization from node A, as described above, but also has an internal clock that it can alternatively use as a source of the synchronization reference signal. Each of them is assigned a relative priority, which indicates the order of preference when using these possible sources of synchronization. So, for example, the synchronization reference signal of the highest priority of node B comes from node A, and its reference synchronization signal of the next after the highest priority (in this case, priority "2") comes from the internal clock of node B.

The fourth column 307 in the table indicates the sequence of nodes through which the synchronization reference signal can be traced. For example, node A is a PRC, so there are no nodes through which its synchronization reference signal is delivered. For node B, the synchronization reference signal is traced to node A when the highest priority reference signal (i.e. provided by node A) is used, and alternatively is traced to B itself when its own internal clock B is used. The brackets are used in the fourth column to indicate when the node is part of the synchronization chain along the packet chain, but only forwards the timing packets without processing or expanding them.

The fifth column 309 indicates the status of the corresponding synchronization reference signal for the network element. Possible conditions include: "PRC network (G.811)"; "locked (G.811)"; "pending (G.812)"; and “pending (G.813),” the latter being a lower-quality clock signal than specified in G.812. These conditions are well known in the art and are defined, for example, in ITU-T Recommendation G.811 (2/97), "Timing characteristics of Primary Reference Clocks"; and ITU-T Recommendation G.812 (2/97), Timing characteristics of Slave Clocks; and ITU-T G.813 (8/96), "Timing characteristics of SDH Equipment Slave-Clocks (SEC)" recommendations, which are all hereby incorporated by reference in their entirety.

The sixth column 311 indicates whether the circuit has a packet-based part, or whether it is carried by synchronous connections. The seventh column 313 in the table shows the result of the observation, for example: maximum time interval error (MTIE), time deviation (TDEV), frequency deviation (FDEV) and others. In the case of a packet-based connection, this may be a PDV measurement. PDV measurements are considered part of the G.8261 standard. These observation results are the result of usually periodic observation, and are installed in the usual manner on a network element performing test observations. These results can be read by the operator to control the quality of the synchronization network, but can also be used by the synchronization network in automatic mode to reconfigure the synchronization network (for example, if the connection shows poor quality). The observation results of MTIE and TDEV are well known in the art and are given in ITU-T Recommendation G.810 (5/96), "Definitions and terminology for synchronization networks", which does not require a more detailed description. The result of the SSM observation is also well known and defined in the aforementioned document ETSI EG 201 783. The frequency deviation is also a well-known concept and does not require a specific definition. The eighth column 315 indicates the functions of the node, for nodes along the chain, which may affect the transmission characteristics of synchronization. This allows you to include in the table some of the transmission characteristics of the synchronization of parts of chains using packet-based transmission.

Fig 4. Characteristics of the transmission of synchronization for part based on the packet

The main difference between physical layer-based communication and packet-based communication (for example, timing carried by IEEE 1588 (PTP) packets) is that in the first case, synchronization is distributed and regenerated by all network elements that are part of the synchronization network. In the latter case, the timing is distributed from the master to the slave, and intermediate nodes may or may not be required to effect the timing (for example, through regeneration). Another main difference is that from the point of view of the slave element, the pending reference signal can be set based on the leading clock generators, and not on the basis of connections, as in the case of traditional methods of synchronization at the physical level.

This means that different information is exchanged between nodes, for parts based on circuit packages, and at the stage of circuit comparison it is necessary to be able to take this new information into account. In some embodiments, a single universal administration protocol running in the control plane can simultaneously operate with different types of synchronization networks (e.g., SyncE and IEEE 1588), although in principle, different protocols can collect different types of information for parts based on packets and synchronous parts , optionally stored in different places, and the circuit comparison unit accesses both places in order to be able to compare different types.

The following packet-based transmission cases can be identified:

a fully packet-based network acting on a packet-based network with a connection;

a completely packet-based network operating over a packet network without a connection;

mixed-layer network of physical layer / packets.

The packet-based method can also be implemented with or without network timing support (for example, timing packets are regenerated or modified at all or some of the network nodes along the chain).

The connectionless scenario will not be described in detail since delays are less predictable. To make delays more predictable, you can use some QoS mechanism on predefined routes. The nodes that are part of the packet-based synchronization network parts can, for example, be of the following types (other types can be envisaged):

nodes only forwarding the timing packet (and thus adding some packet jitter);

time regenerating nodes (for example, the IEEE 1588 edge clock); this is much like a clock based on the physical layer (for example, in relation to pending clocks, clock quality, etc.), see FIG. 4, first chain leading to F;

nodes that expand the packet data (for example, transparent nodes of the IEEE 1588 clock) to expand the data to indicate the delay values defined on the node, see, for example, FIG. 4, the chain leading to H; and

nodes specified by the routing protocol as an alternative route in case of network failures (in this case, some interaction with the routing protocol data (for example, LSA OSPF) can be used to obtain a picture of the available routing routes.

These different types of nodes can affect the timing of the synchronization information carried by the packets, and therefore the information about these nodes may represent transmission characteristics of the synchronization circuits. The comparison and selection of the mechanism chains in this context can be expanded to take into account, for example, transmission characteristics, for example, the number of nodes in the chain, the number of nodes that do not provide timing support, the number of boundary nodes of the clock generator, the number of transparent nodes of the clock generator, etc.

In the case of methods on a mixed basis of the physical layer / packets, this information can also be added by each node. The comparison and selection mechanism can also take into account other factors, for example, the reliability of one method compared to another (in particular, the method based on the physical layer is not affected by changes in packet delay and can be considered as having first priority).

Therefore, the data to be exchanged between the nodes and the administration network may include the above information.

In order to avoid excessive growth of the tables, the synchronization network can, optionally, be divided into parts, each of which contains only data for the corresponding section of the entire synchronization network.

FIG. 5. Stages using synchronization transmission characteristics

In FIG. 5 shows a flow diagram of steps for configuring a synchronization network according to an embodiment. At step 200, the node is added to the hybrid synchronization network. Step 210 involves starting the configuration of the node by communicating with other nodes to exchange information to determine the circuits leading to the node. This may include, for example, obtaining a copy of the table shown in FIG. 4, from other nodes, or filling out rows of such a table by requesting information from other nodes. At step 220, source information is determined for each circuit, and at step 230, synchronization transmission characteristics are determined. This may include, for example, obtaining information from other nodes, or from a central location, to populate the second and eighth columns of the table shown in FIG. 4, for lines corresponding to the desired node. Other methods may be provided for obtaining or storing such information.

At step 240, the chains leading to the node are compared using source information and timing characteristics to select the chain to be used. Some methods for implementing this step are described in more detail below with reference to FIG. 6-9. Optionally, a new selection is recorded, optionally, in the table's priority column, and if the second and third priority levels, etc., are defined for the remaining circuits, these priorities can also be recorded. This allows the node, if necessary, to instantly switch to the backup circuit, without a complete revision of priorities.

Step 250 shows the normal operation of the synchronization network to provide a synchronization reference signal for the corresponding node of the communication network, as shown in FIG. 2, for example, based on synchronization information passed from a remote synchronization source along a selected circuit. At 260, the circuit information is periodically updated, possibly as a result of a failure or alarm condition, or reconfiguration of any part of the communication network. This can be done by exchanging information between nodes, which is described in more detail below.

Updating a database table without creating timing cycles

Now we will consider how the table contents changes in the event of a failure in the synchronization network. Assume that a failure has occurred (for example, a fault in the cable), as a result of which the synchronization connection between nodes F and G has broken. This means that node G will have to receive synchronization from another source, and in order to avoid timing cycles, other nodes may also need to be reconfigured . This process can be coordinated using a protocol for distributing table updates so that each node (including the one connected to the integrated synchronization network administration network 201) has complete information about the main reconfiguration decisions. Information changes fragmentarily, as a result of which several updated tables are created and sequentially distributed.

For example, upon detecting the loss of its synchronization source, node G will check its own table and find out that it cannot immediately receive its synchronization reference signal from node H, since this will lead to the creation of a timing cycle (the table indicates that the source of node H is traced to the node G) Then, the node G selects, in the alternative, the reception of its reference clock signal from its own internal clock. The node G updates the table 300 to reflect this change, and distributes it to other nodes in the synchronization network. When distributing a table from node to node, changes are made gradually to adjust previous changes. In particular, if an observation result indicating “alarm” is generated on node G, for example, as a result of an observer detecting a malfunction, for example, frame loss or an MTI exceeding a certain threshold, this “alarm”, in turn, causes a transition connections of the highest priority of the node G from the state "blocked" to the state of "failure". In response to this failure, the source of the synchronization reference signal with the third after the highest priority of the node G switches from the "standby" mode to the "locked" state (in the sense that it is used by the node G). This is emphasized by the fact that in the algorithm for selecting a reference signal, the condition for preventing timing cycles is more significant than the choice of a reference signal with the highest priority. An additional consequence is that the H node changes, demonstrating the quality transition from the G.811 quality reference signal to the G.813 quality reference signal.

FIG. 6-8. Circuit comparison

In FIG. 6 shows some of the steps provided for in one possible implementation of step 240 of FIG. 5. As shown, in the first step 310, it is determined which nodes are in parts of the chains using packet-based transmission. At 320, it is verified that none of the circuit paths uses the same node twice to avoid timing cycles. At 330, the number of nodes in each of the chains is determined, and this number is used as a measure of transmission quality. In this case, for example, column 4 in FIG. 4 to determine which nodes are passed and use the eighth column in FIG. 4 to determine which of these nodes are in the package-based part.

At step 340, the chains are compared to determine if one circuit has significantly fewer nodes. The threshold can be applied, depending on, for example, how many chains are compared, or depending on the length of the chains. If the shortest chain has, say, less than 10 nodes, it will be selected, but if the shortest chain has 20% fewer nodes than the chain, the second after the shortest, you can select it. Many other similar algorithms can be envisaged. The comparison may include, as a factor, the characteristics of the source if the sources differ in quality, such as jitter or reliability. If no circuit is clearly optimal, then at 350, additional synchronization transmission characteristics can be determined for use in comparison.

In FIG. 7 shows a step 352 for determining the speed of connections between nodes. This is not directly shown in the table in FIG. 4, but can be determined in various ways, by querying either a central database or local storage on a node indicating the network topology and specific values of the data rate for connections between nodes. The request may indicate connections based on which nodes participate. Since in step 330 it is possible to determine which nodes are passed, for example, using column 4 in FIG. 4, and use the eighth column in FIG. 4 to determine which of these nodes are in the package-based part.

At 354, delays and / or jitter can be determined and accumulated for each of the chains based on connection speed information. At 356, the chains can be compared based on their accumulated totals to determine if one chain is really much better than the others. Again, you can use the appropriate thresholds, as discussed above, to determine to what extent or how much the chain must outperform its competitors in order to be chosen. At step 358, the selected circuit is output, or, if the best circuit does not meet the threshold, then additional characteristics are determined.

For example, at 362, a determination is made of how many nodes in each circuit are involved in expanding the synchronization information. This can be done based on the node function information in the eighth column of FIG. 4, described above, to determine if there are nodes that expand, for example, by measuring the actual jitter or delays along the chain, in order to make it possible to compensate or at least assume these values to be replaced with actual values. At block 364, the chains are again compared to determine if one chain is really much better than the others. Again, you can choose the appropriate thresholds. If the best circuit meets the thresholds, it is selected, and the corresponding output signal is supplied to the circuit selection unit, otherwise, additional characteristics can be determined at 366.

In FIG. 8 shows a step 372 of determining any nodes that regenerate synchronization information, for example, boundary nodes of a clock. Again, this can be done based on the node function information in the eighth column of FIG. 4 described above. At block 374, the chains are again compared to determine if one chain is really much better than the others. Again, you can choose the appropriate thresholds. If the best circuit meets the thresholds, it is selected, and the corresponding output signal is sent to the circuit selection unit; otherwise, additional characteristics can be determined at step 376. Other transmission characteristics, such as reliability or other factors, may be such other characteristics.

FIG. 9. Comparison involving the integration of synchronous parts and parts based on packages

In FIG. 9 shows a sequence of steps similar to those shown in FIG. 6, but with the additional step of combining synchronous parts and parts based on packages. As in FIG. 6, some steps are provided for in one possible implementation of step 240 of FIG. 5. As shown, in a first step 310, it is determined which nodes are in parts of the chains using packet-based transmission. At 320, it is verified that none of the circuit paths uses the same node twice to avoid timing cycles. At 330, the number of nodes in each of the chains is determined, and this number is used as a measure of transmission quality. In this case, for example, column 4 in FIG. 4 to determine which nodes are passed and use the eighth column in FIG. 4 to determine which of these nodes are in the package-based part. At step 335, delays (or other characteristics, such as jitter) of the packet-based parts of each circuit are combined with the accumulated delays of any synchronous parts. This can be done based on the number of nodes or other measures of delay or jitter.

At 345, the chains are compared to determine if one circuit has significantly higher transmission quality (for example, fewer nodes). The threshold can be applied, depending on, for example, how many chains are compared, or depending on the length of the chains. If the shortest chain has, say, less than 10 nodes, it will be selected, but if the shortest chain has 20% fewer nodes than the chain, the second after the shortest, you can select it. Many other similar algorithms can be envisaged. If no circuit is clearly optimal, then at 350, additional timing transmission characteristics can be determined for use in comparison.

FIG. 10. Centralized circuit comparison

In FIG. 10 is an example showing features similar to those shown in FIG. 2, but characterized in that the functions related to the configuration, for example, circuit comparison, are carried out centrally. The figure shows a centralized synchronization network administration system 500 having a circuit information administration unit 420, a circuit information database 435, and a circuit comparison unit 430.

There is a node 410 synchronization network, which outputs a reference synchronization signal or a clock signal for use by synchronous parts of the communication network. This node of the synchronization network has a circuit selection unit 440 for selecting which of the various synchronization chains leading to this node is selected to provide synchronization information. Similarly to FIG. 2, this timing information is used by the timing reference signal generator 450 in a conventional manner to generate a clock signal. The circuit selection unit is controlled by the output of the circuit comparison unit 430 from the centralized synchronization network administration system.

The circuit comparison unit uses information about the source and the characteristics of the synchronization transmission stored in the database 435. The block information administration unit 420 fills and maintains this database, receiving information from all nodes in the synchronization network. The advantage of centralizing these functions, compared to distributing them across all nodes, is that it allows you to reduce delays when distributing information over large networks. In addition, the centralized version simplifies operator intervention. In principle, it is possible to maintain a centralized database of circuit information, and the circuit comparing block algorithm can be performed locally on each node. This will allow you to distribute the need for computing resources, but to increase the need for communication resources between nodes and a centralized database.

FIG. 11. an example of selecting circuits in a mixed synchronization environment based on the physical layer / packets

In FIG. 11 shows a hybrid synchronization network similar to that shown in FIG. 1 (but not identical, for example, the node G is, according to FIG. 11, a transparent node of the clock generator). In FIG. 11 shows two chains leading to node N. The first chain, indicated by a dashed line, passes from O through F, I, H, and M. The second chain, indicated by a solid line, passes from E through F, G, H, and M. Both chains have the same number of nodes, and they cannot be distinguished only by these criteria, as is done in the steps shown in FIG. 6. However, the chains can be distinguished using step 362 of FIG. 7. The second chain provides better timing support as it passes through node G, which is a transparent clock node, and not through node I, which does not process packets at all. This will be reflected in the last column of the table in FIG. 4 if FIG. 4 is modified in accordance with the network shown in FIG. eleven.

Expansion to time / phase synchronization

The use of time (and / or phase) synchronization usually involves the presence of a synchronization network using a packet-based method. In some cases, it is possible to propose a solution according to which the packet-based method receives support from the underlying physical layer method, which is used to distribute frequency synchronization (this is sometimes called “route support”, OPS).

In addition to the need for a temporary (or phase) reference signal, the main difference between the packet-based method used for frequency distribution and the method that also distributes time consists in the need for a two-way protocol. Thus, timing packets are transmitted from the master to the slave and from the slave to the master. Another characteristic is that in order to deliver accurate time, in many significant network scenarios, support from network nodes (for example, IEEE 1588 edge clocks) will be required.

Based on the fact that a two-way protocol is established for the downlink timing stream distributed over the same nodes as the corresponding uplink stream, this is a direct extension of the considerations presented in the previous sections.

In the case of a time synchronization network that supports the physical layer synchronization method, additional information about the presence of OPS is considered. In this case, the node must simultaneously support two synchronization databases: one for distributing frequency synchronization and the other for distributing time synchronization.

A distributed approach is preferred because it can guarantee the best performance in terms of convergence time.

FIG. 12, 13. Extension of existing control plane protocols

In some embodiments, OSPF expands in a similar fashion to the optical plane's control plane (e.g., ASON / WSON). In the case of ASON, OSPF was expanded to G.OSPF-TE (“Generalized OSPF with Traffic Engineering” extension) to collect / disseminate technology-specific information, such as wavelength continuity, physical damage, etc.

In this example, OSPF expands by defining an LSA (Connection Status Alert) of a TE connection that can also carry synchronization information. In particular, for the synchronization applications discussed here, G.OSPF-TE is adapted to distribute the topological information of the synchronization network. Note that no extension of the signaling protocol is required, since the information distributed through the advanced OSPF is sufficient for the nodes to correctly establish the synchronization routes.

Each TE connection in an OSPF network describes, using LSA, a TE connection (according to RFC 3630 - Extending Traffic Engineering (TE) to OSPF version 2) containing a TLV (type-length-value) connection. TE LSAs contain a certain number of subobjects called sub-TLVs. Each of these subobjects describes some of the characteristics of the TE connection.

Each TE connection contains one or more connection components (LCs) representing a physical connection between two adjacent nodes. The first parameter associated with the LC is the local identifier that can uniquely identify it: the component ID of the connection. Other parameters are technology dependent and, in this particular context, relate to synchronization.

In FIG. 12 illustrates an example of a sub-TLV synchronization specification that contains two connection components and associated synchronization parameters. As the symbol ^(*) on the top line indicates, this value of 1 indicates that this is only a sync extension. In the second line, the symbol ^(**) after the value of the sub-subtype serves to indicate that others can be added for other synchronization methods (for example, based on packets, SDH, etc.), and the set of parameters will change accordingly.

An additional set of synchronization parameters is associated with the node (for example, the quality and state of the clock generator). These parameters can be propagated through a particular sub-TLV in the LSA of the TE connection, as shown by way of example in FIG. 13.

In FIG. 13 shows an example of synchronization of a specific sub-TLV (case of a node), in which, again, in the first line there is an indication ^(*) that this is only an extension of synchronization. In the second line, the symbol ^(**) is used to indicate that this is the only type of node. Alternatively, you can add other sub-subtypes (with increasing id, i.e. 3, 4, etc.) for other types of nodes, and the set of parameters will change accordingly. In the fourth row, the symbol ^(***) is used to indicate that this node function may correspond to the contents of the eighth column in FIG. 4, for example, PRC networks, G.8263 TC, etc.

The second parameter in the sixth row in this example is the observation status, which corresponds to the contents of the seventh column in FIG. 4. Thus, it can be seen how this can be used to exchange information between nodes to populate the table in FIG. 4, which can then be used as the basis for various circuit comparison algorithms, to select the best synchronization circuit for each node and, thus, configure the synchronization network.

In addition to OSPF, other existing or new protocols can be considered, provided that they provide additional functionality. For example, LMP (Connection Administration Protocol), which runs on a pair of nodes and is designed to administer TE connections, can be used to extend the synchronization solution.

Other symptoms

A new protocol adapted to transfer new transmission characteristics can be implemented with the possibility of coexistence with existing synchronization mechanisms. If any conflict occurs, it may be necessary for the nodes to ignore some parts, for example, the “SSM” protocol, which may need to be ignored in order to avoid possible inconsistencies.

For the case of synchronization networks based on the physical layer (for example, SyncE, SDH), all of the protocol packets described above are distributed in a single IP hop. This means that all network elements participating in the synchronization network (or being its edge nodes) must be properly addressed (the protocol must operate on the basis of connections). In other words, packets should not be transported transparently through network elements that are part of a synchronization network.

Other changes and embodiments may be envisaged within the scope of the claims.

Claims (16)

1. A method for configuring a synchronization network node, the method comprising the steps of:
a) determining information about the synchronization sources of multiple synchronization circuits for transmitting synchronization information from the synchronization source to the node to provide a reference synchronization signal, and part of at least one of the chains uses packet-based communication to transmit synchronization information,
b) automatically determine the transmission characteristics of the synchronization of the part that uses packet-based communication, and
c) automatically compare the circuits leading to the node, using certain source information and using certain characteristics of the synchronization transmission, to select which of these chains to use to provide the reference synchronization signal for the node. 2. The method according to claim 1, wherein the step of determining the transmission characteristics of the synchronization comprises determining information about the nodes used to transmit packets used for packet-based transmission. 3. The method according to claim 2, in which the step of determining information about the nodes comprises determining information related to the expansion of synchronization information carried by packets by any node. 4. The method according to claim 3, in which the step of determining information related to the extension, comprises determining information related to the regeneration of synchronization information on another node along the chain. 5. The method according to claim 3, in which the extension contains information about delays in relation to packets carrying synchronization information, the delays being determined by another node along the chain. 6. The method according to claim 5, in which the information about the delays contains any one or more of the delays internal to the node, delays between nodes and the magnitude of the change in delay. 7. The method according to any one of the preceding paragraphs, having the step of determining information about the synchronous parts of any circuits leading to the node, and using this information in the comparison. 8. The method according to any one of claims 1 to 6, in which the step of determining the characteristics of the synchronization transmission is carried out by the node, the node having a record of synchronization circuits with itself and with other nodes. 9. The method according to any one of claims 1 to 6, having the steps of receiving, on the update node, synchronization transmission characteristics from a neighboring node, reviewing the selection based on the updated synchronization transmission characteristics, selecting a new one of the synchronization circuits to pass synchronization information for use as a reference signal synchronization and sending updates to neighboring nodes to indicate the updated characteristics of the transmission synchronization and a new choice of circuit. 10. A node for a synchronization network, wherein the node has
a) a circuit information administration unit, configured to determine information about the sources of a plurality of synchronization circuits leading to a node, for passing synchronization information from synchronization sources to a node, to provide a synchronization reference signal, wherein part of at least one of the chains uses communication based on packets for transmission of synchronization information, and the circuit administration unit is also configured to automatically determine the transmission characteristics of synchronization ty, which uses the transfer of packet-based, and
b) a comparison unit, configured to compare the circuits leading to the node, using the characteristics of the synchronization transmission, to select which of the chains to use to provide the reference synchronization signal for the node. 11. The node of claim 10, in which the characteristics of the synchronization transmission, determined by the circuit administration unit, contain information related to the expansion of the synchronization information carried by the packets by any node. 12. The node of claim 10, having a part of a control plane for controlling a communication network synchronized by a synchronization network, wherein the circuit administration unit and the comparison unit are implemented by a part of the control plane. 13. The administration system for the synchronization network, and the administration system is configured to implement the configuration method according to any one of claims 1 to 9. 14. A synchronization network having a number of nodes and having a number of synchronization chains for transmitting synchronization information from a synchronization source to nodes of a hybrid synchronization network, the hybrid synchronization network having an administration system configured to implement the configuration method according to any one of claims 1 to 9 . 15. A computer-readable medium having a computer program stored on it, and the computer program contains instructions that, when executed by the processor, instruct the processor to implement the configuration method according to any one of claims 1 to 9. 16. A database for configuring a node of a synchronization network, the database comprising recording synchronization sources of a plurality of synchronization circuits for transmitting synchronization information from synchronization sources to a node to provide a synchronization reference signal, wherein part of at least one of the chains uses transmission based on packets, and the database also contains a record of transmission characteristics of the synchronization of the part using packet-based communication, and a record of which chain is selected to provide a reference drove synchronization.

Images